gel-casting; α-alumina;
acrylamide; drying;
green machining;
de-binding;
sintering.
A B S T R A C T
Key Words:
(*)To whom correspondence should be addressed.E-mail: [email protected]
INTRODUCTION
Ceramic parts are formed by consoli-dating powders into the desired andoften rather complex shapes and den-sifying them at higher temperatures.The main challenge in this process isto achieve parts with defect-freemicrostructures because pores, inclu-
sions, or any other inhomogeneity ofthe microstructure cause stress con-centration during loading of a part,and subsequently the risk of failure.As more advanced ceramics arebeing used in industry [1], it is a greatchallenge to fabricate such complex-
Forming Alumina Parts Using AcrylamideGels
Abolfazl Barati1, Mehrdad Kokabi*, and Mohamad Hossein Navid Famili
Polymer Engineering Group, Tarbiat Modaress University (TMU),
P.O. Box: 14115-111, Tehran, I.R. Iran
Received 20 July 2002; accepted 14 December 2002
Today, the fabrication of complex-shaped ceramic parts in large quantities has beenlimited to the processes required extensive and expensive machining (i.e., injec-tion moulding or pressure slip-casting). These technologies for shaping ceramics,
however, possess several shortcomings that limit their use as a complex-shape formingmethod. Because of these limitations, “gel-casting” which was lacking in above deficien-cies, has been proposed as a general and desired method for shaping any ceramic pow-der. In this work, gel-casting of α-alumina is investigated based on in situ polymerizationand gelation of acrylamide monomers as the setting mechanism for forming the greenpart. Special attention was paid to the preparation of fluid, castable slurries with a highloading level of α-Al2O4 (>80 wt.%) and low viscosity (0.65 Pa.s) by using polyelectrolytedispersants (i.e., poly(methacrylic acid) salts), also gelation of acrylamide (effect of gela-tion factors on idle time), mould materials, drying of parts through the liquid desiccant dry-ing method and its comparison with conventional drying technique, green machining,debinding and sintering.
Iranian Polymer Journal, 1122 (2), 2003, 127-138
shaped parts with high reliability at acceptable costs.An ongoing focus is the development of near-net shapefabrication processes that can produce complex-shapedparts with a minimum of machining by optimizing andcontrolling each fabrication process step in order tominimize the number and size of microstructuraldefects to within design limits [2].
The existing forming processes, such as slip-cast-ing, pressing and injection moulding possess severalshortcomings that limit their use as a complex-shapeforming methods and do not conform in all of the afore-mentioned requirements [3-5].
As such, forming of the parts by using organic gels,gel-casting is a net-shape forming process similar to
the other near-net shape forming processes (i.e., pres-sure slip-casting, injection moulding), but with severaladvantages [6]. Compared to slip-casting, gel-castingresults in much more homogeneous material with nodensity differences over the parts. Compared to injec-tion moulding, gel-casting does not need expensivemoulds, uses only small quantities of organic binder(typically 4 wt.%), uses water or other low viscositysolvent as the suspension media and does not includethe critical binder removal step [7].
The advantages of gel-casting can be summarized
as, (a) capability of producing complex parts, (b) easeof implementation owing to its similarity with othersuspension forming technique, (c) low equipment cost,(d) low cost mould material, (e) high green strength, (f)excellent green machinability, (g) very homogeneousmaterial properties and (h) ease of binder removal com-pared to injection moulding [7].
Other techniques like freeze-casting [8], aqueousinjection moulding [9], direct coagulation-casting(DCC) [10] and hydrolysis-assisted solidification [11]are gaining renewed attention by the success of gel-casting.
In our group the development of gel-casting, basedon the research results published by the Young et al.[12] started 10 years ago, but the first article on gel-casting of engineering ceramics, mainly reaction bond-ed silicon, was published a few years later [13]. In thiscontext, the aim of this paper is to investigate the gel-casting of Al2O3 powders; preparation of fluid, castableslurries with a high Al2O3 loading level, gelation ofacrylamide as a polymeric binder and the parameterseffect on it, choosing the mould material, drying ofparts through the liquid desiccant drying method andcomparing it with conventional drying techniques,green machining, de-binding and sintering.
128
Forming Alumina Parts Using Acrylamide Gels Barati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003)
Table 1. Composition of samples.
Material Function Chemical formula Characteristics Supplier
Fused α-alumina
Acrylamide (AAm)
N,N'-Methylene bis acrylamide (MBAM)
Ammonium persulphate (APS)
N,N,N',N'-Tetra methylethyl diamine
(TEMED)
Poly(ethylene glycol) (PEG)
Sodium salt of poly(methacrylic acid)
Ammonium salt of poly(methacrylic acid)
Ceramic powder
Monofunctional
monomer
Bifunctional monomer
(cross-linker)
Initiator
Accelerator (catalyst)
Liquid desiccant
Dispersant
Dispersant
Al2O3
C2H3CONH2
(C2H3CONH2)2CH2
(NH4)2S2O8
C6H16N2
HO(C2H4O)nH
(C4H5O2- Na+)n
(C4H5O2-N+H4)n
Figure 1 (a to c)
Mw 71.08
Soild white powder
Mw 154.2
Solid white powder
Mw 228.2
Solid white powder
Mw 116.2
Liquid, yellow
Ave. Mw 950-1050
Ave. Mw 15000
Aqueous solution
40 wt.%
Ave. Mw 15000
Aqueous solution
40 wt %
Good Fellowa
Merckb
Sigmac
Merckb
Merckb
Merckb
R.T. Vanderbiltd
R.T. Vanderbiltd
(a) Cambridge Science Park, Cambrige CB44DJ, England; (b) E. Merck, D6100 Darmstadt, Germany;
(c) Sigma Chem. Co, USA; (d) R.T. Vanderbilt Co., USA. (Figure 1)
Forming Alumina Parts Using Acrylamide GelsBarati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003) 129
EXPERIMENTAL
MaterialsThe specification of the materials used in this work aresummarized in Table 1 and Figure 1.
InstrumentsInstruments which were used in this study are as fol-lows:
- Brookfield viscometer (Shanghai balance instru-ment factory, model NDJ-4, type LVF).
- One axis tools machine (Tabriz, T N50BR).- Scanning electron microscopy (Philips XL30,
magnifying to 105).- Programmable furnace with computer controlling,
(Exciton, model Azar 1500).
(a)
(b)
(c)
0 10 20 40 50 60 70 80( 2T )Alum - RA
0
20
40
60
80
100
( % )
Sample identification . : Alum - Ra
30
Figure 1. (a) Shape, (b) particle size distribution and (c) x-ray
diffraction of α-alumina powder.
Organic
monomer
Ceramic
powder
Water
dispersants
Initiator
accelerator
Slurry
Casting
Drying
Demoulding
Binder out
Sintering
Machining
Figure 2. General gel-casting process flowchart.
Pre-mixAAm
(wt %)
MBAM
(wt %)AAm / MBAM Viscosity (mPa.s)
A1
A2
A3
A4
A5
A6
5
6
7
8
9
10
0.4
0.2
0.2
0.2
0.2
0.2
12.5
30
35
40
45
50
0.9
0.9
1.0
1.0
1.1
1.2
Table 2. Composition and viscosity of aqueous mixtures of
binders (pre-mixes).
130
Barati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003)
Forming Alumina Parts Using Acrylamide Gels
- Ultrasonic shaker (Bransonic model B200, outputfrequency: 40 kHz).
- pH meter (Eutech instruments, type WP2).ProcedureSlurry Preparation and CastingThe gel-cast process flowchart typically looks like theother ceramic processes especially slip-casting, ( 2) [12].Firstly, by using magnetic or mechanical stirrer (withturbine blade), pre-mix solution that containsmonomers, solvent and dispersant is prepared. Thecomposition of the pre-mixes and their viscosity aregiven in Table 2. After adding the dispersant to the pre-
mix solution, alumina powders with various loadinglevels were added to solution and stirred by usingmechanical stirrer for at least 2 h to promote dispersion.
The resulting slurry was degassed by ultrasonicshaker. Vacuum was also applied until no further airbubble was released. Then accelerator was added andgently agitated for at least 15 min. A 10 wt% aqueoussolution of ammonium persulphate, i.e. the initiator,was added to slurry and manually stirred for 1 min.After that the slurry was cast into the mould at roomtemperature. A combination of epoxy and steel mouldswere used in this work. A non-silicone based mouldrelease agent (dioctyl phthalate) was also employed.Gelation took place in 3 to 75 min after casting,depending on the processing conditions. After comple-tion of polymerization process, the ceramic green partwas demoulded and dried through liquid desiccant
TEMED (mL) / 100g of powder
0 0.05 0.1 0.15
3000
2500
2000
1500
1000
500
0
Figure 3. Effect of TEMED as dispersant on the viscosity of
35 vol% alumina slurry.
0 0.1 0.2 0.3 0.4 0.50
1000
2000
3000
4000
5000
6000
ammonium salt
sodium salt
PMAA salt (mL) / 100 g of powderFigure 4. Effect of ammonia and sodium salts of PMAA
on the viscosity of 45 vol % alumina slurry.
ammonium salt
sodium salt
Spindle speed (rpm)
1200
1000
800
600
400
200
00.1 1 10 100
Figure 5. Comparison between viscosity reduction of poly-
electrolytes on 40 vol% alumina slurry in 0.3 mL of polyelec-
trolyte per 100 g of powder.
IdIe time
0 100 200 300 400 500 600
40
30
20
10
0
Time (s)
Figure 6. Change of pre-mix temperature during gelation
(typical pre-mix which has 7.5 vol% of acrylamide and molar
ratio of AAm per MBAM equal to 10).
Forming Alumina Parts Using Acrylamide GelsBarati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003) 131
method by immersing it into 20 wt% of aqueous ornon-aqueous solution of PEG1000. After 4 h, greenpart, which lost about 25-30 wt% of its water content,was removed from liquid desiccant and drying fol-lowed in an oven at 60-80°C. Once dried, the gel-castgreen bodies are sufficiently strong for machiningoperations. After machining, polymer was burnt outunder controlled heating program, which was deducedfrom thermogravimetric analysis (STA) of acrylamidegels. Subsequently sintering of the part was carried outin air. A typical sintering scheme was 0.1 C/min to600 C and 2 C/min from 600 to 1450 C.
Measuring of Physical and Chemical PropertiesParticle size and particle size distribution of aluminapowder were measured by scanning electronmicroscopy (Figure 1). Viscosity of pre-mix solutionand slurry with and without dispersing agent weremeasured by Brookfield type viscometer, using spindleNo. 1 (for pre-mix) and No. 2 (for slurry) at angularvelocity of 3 rpm in water bath at 25 C. The variationsof viscosity versus shear rate were monitored. All
measurements were repeated to check for reproducibil-ity and averaged.
The bulk densities of complex-shaped gel-castparts after sintering were determined by Archimedesmethod.
RESULTS AND DISCUSSION
Solid Loading of the SlurryIn gel-casting, the slurry is cast in a non-porous mouldwherein it is subsequently set by the gelation reaction.Accordingly, the green density of the part formed cor-responds to the solid loading of the slurry. Hence, it isbeneficial to obtain a solid loading as high as possiblein order to achieve higher dimensional tolerances byminimizing the possible distortion during drying andsintering shrinkage. On the other hand, the shaping ofintricate parts requires the slurry to easily fill moulds ofpossible complex geometry. One way of achieving thisis to use pre-mix solution of low viscosity. As shown inTable 2, the pre-mixes A1 to A6 (which contain 5 to 10
Exp. noMonomer concentra-
tion (wt %)Molar ratio of monomer
to cross-linkerMolar ratio of initiator to
total monomers Temperature (oC)Ceramic loading
level (vol %)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
6
7
8
9
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
22
22
22
22
16
18
20
22
22
22
22
22
22
22
22
22
22
22
22
22
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.3
0.4
0.5
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
25
25
25
25
25
25
25
25
25
25
25
25
15
25
35
45
25
25
25
25
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20
30
35
40
Table 3. Experimental design for determining the effects of various factors on the idle time of gelation process.
132
Forming Alumina Parts Using Acrylamide Gels Barati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003)
wt % binder) have viscosities similar to that of purewater (i.e., 0.9 mPa.s).
Using a proper dispersant for the powder is alsoessential to prevent any participation in slurry. Here,three dispersants were tested. These included sodiumand ammonium salts of poly (methacrylic acid) andTEMED. TEMED, Which is used as an accelerator ingelation process, acts as a good dispersant for our pow-der. TEMED With two amine groups is adsorbed on thesurface of the powder, prevents strong hydrogen bond-ing between water and alumina by making poor hydro-gen bonding with water.
As shown in Figure 3, viscosity of the slurry wassignificantly decreased by adding small amount ofTEMED. Concentration of TEMED in slurry must becarefully controlled until gelation would not occur dur-ing casting in a mould. Sodium and ammonium salts ofpoly (methacrylic acid), which act as polyelectrolytes,are adsorbed on ceramic powder surface and preventstheir coagulation [14]. As shown in Figure 4, by usingthese polyelectrolytes, viscosity of the slurry was high-
ly decreased (21 times). The minimum viscosity isobtained at about 0.3 mL of both salts per 100 g of drypowder. Higher concentration of polyelectrolytesresults in an increase in the slurry’s viscosity. As shownin Figures 3 and 4, addition amount of dispersants toslurry has a limiting value. At this level and above it,the particle surface is saturated with the adsorbed dis-persant and that the viscosity of solution is no longerdominated by the concentration of dispersant. Also,excess amount of polyelectrolyte in solution results infree polymer, which increases the viscosity of slurry.Figure 5, compares performance of the polyelectrolyteson viscosity reduction of alumina slurry. The sodiumsalt was better than the others but in the processing ofadvanced ceramics, ammonium salt is commonly pre-ferred because the sodium salt can lead to the formationof undesirable liquid phase during sintering.
Gelation ProcessIn this study, polymerization occurs under free-radicalmechanism, which is an exothermic reaction. Thus, the
2000
1600
1200
800
400
05 6 7 8 9 10
y = 224993 X_ 2.7423
R2 = 0.9734
Vol(%) of AAm in pre-mix
Figure 7. Effect of acrylamide concentration on idle time
of gelation of pre-mix solution.
y = 160.5 X _ 1776.5
R2 = 0.9734
Molar ratio of AAm / MBAM
2000
1600
1200
800
400
014 16 18 20 22 24
Figure 8. Effect of acrylamide concentration on idle time
of gelation of pre-mix solution.
y = 5E+06 X_ 2.5562
R2 = 0.9673
6000
5000
4000
3000
2000
1000
00 5 10 15 20 25 30 35 40 45 50
Temperature(OC)
Figure 10. Effect of temperature of solution on idle time of
gelation of pre-mix solution.
y = 84.639 X_ 1776.5
R2 = 0.9766
Molar ratio of initiator to total monomer
0
400
800
1200
1600
2000
0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55
Figure 9. Effect of initiator concentration on idle
time of gelation of pre-mix solution.
Forming Alumina Parts Using Acrylamide GelsBarati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003) 133
initiation of polymerization in pre-mix can be deter-mined by changes in the temperature of solution. Theidle time, tidle, i.e. the time between the addition of theinitiator and the beginning of polymerization can bemonitored. This is equivalent to the time available forcasting the slurry during processing. Figure 6, showsthe changes in temperature of a pre-mix solution duringgelation and indicates the idle time.
Idle time is affected by the factors such asmonomer concentration, monomer to cross-linker ratio,initiator concentration, amount of ceramic loadinglevel in slurry and solution temperature, (Table 3). Fig-ures 7-11 show the variation of idle time with any ofthe above factors. As shown in those Figures, idle timedecreases with increase of monomer concentration, ini-tiator concentration, amount of ceramic loading level
Exp.
no.
1
2
3
4
5
6
7
8
9
10
11
12
Alumina
Alumina
Alumina
Alumina
Alumina
Alumina
Alumina
Alumina
Alumina
Alumina
Alumina
Silicon
30
30
4
4
4
4
4
4
4
4
4
5
5
5
5
7.5
5
5
5
5
5
5
5
5
40
20
40
40
40
40
40
40
40
40
40
40
None
None
None
None
Glycerol
(10%)
Glycerol
(20%)
Glycerol
(30%)
Glycerol
(10%)
Glycerol
(10%)
Glycerol
(20%)
Glycerol
(20%)
Glycerol
(20%)
180
180
180
180
180
180
180
180
180
325
180
180
Low
Low
Low
Low
Low
Low
Low
Low
Low
Low
High
Low
0
0
0
0
0
0
0
5
10
0
0
0
Dust
Dust
Small
shavings
Very Small
shavings
Mostly dust
Small
shavings
Large
shavings
Small
shavings
Large
shavings
Dust
Large
shavings
Small
shavings
Small chips
moderate
chips
Small chips
Very small
chips
Small chips
No chips
Plastic
deformation
No chips
Plastic
deformation
Moderate
chips
Huge chips
No chips
Good
Good
Good
Good
Good
Excellent
Excellent
Excellent
Excellent
Excellent
Bad
Excellent
Type
of
cera
mic
Ave
rage
par
ticle
size
(µm
)
Mon
omer
conc
entr
atio
n
(vol
%)
Mol
ar r
atio
of
mon
omer
to
cros
s-lin
ker
Wei
ght
ratio
of
plas
-
ticiz
er t
o m
onom
ers
For
war
ding
vel
ocity
of t
he t
ools
on
the
part
(fe
edin
g ra
te)
Wat
er c
onte
nt o
f
the
part
(w
t %
)
Edg
e an
d co
rner
cond
ition
s
Sur
face
fin
ish
Turn
ings
mor
phol
ogy
Ang
ular
vel
ocity
of
spin
dle
Table 4. Results of gel-cast parts green machining.
134
Forming Alumina Parts Using Acrylamide Gels Barati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003)
and solution temperature, but increases with increase ofmonomer to cross-linker ratio. Effect of monomer tocross-linker ratio and amount of ceramic loading levelon idle time, in the range of experimental conditions,are linear but the other factor affects exponentially.Although, the effect of the solid filler in acceleratingthe gelation (curing) reaction was reported for ther-moset polymers [15-17], but for as-yet undeterminedreasons, for catalytic role of ceramic powder, the idletime is reduced significantly.
Casting Moulds Gel-casting moulds can be fabricated from a widerange of material including metals, glass, petroleumwax and polymeric materials. Since the casting slurriespresent low viscosities, the mould needs to be designedin order to avoid any leakage. Gel-cast parts are weak-er than injection moulded part. Hence, the mouldshould allow to be removed from the parts (vs. the partbeing removed from the mould). In order to make
unmoulding easier, effective mould release agents areneeded (e.g., dioctyl phthalate (DOP), silicon-basedspray or oil, non-silicon based oil, etc.). Some mouldrelease agents will also inhibit the polymerization reac-tion (e.g., silicon-based spray or oil).
With this in mind, we have investigated the use ofseveral mould materials. The compatibility with thegelation reaction and ease of unmoulding has beenassessed qualitatively. We have used different mouldmaterials as metal, Plexiglas and casting epoxy resins.The latter is interesting, as it is cheap and can be usedto fabricate a complex mould starting from, e.g. a pos-
y = -26.76 X +1341
R2 = 0.9977
Vol% of alumina loading in pre-mix0 10 20 30 40 50
900
700
500
300
100
0
Figure 11. Effect of ceramic loading level on idle time of gela-
tion of pre-mix solution.
Non-aqueous desiccantsolution
Aqueous desiccant solution
100
90
80
70
60
50
40
30
20
10
00 20 40 60 80 100 120 140
Time (min)
( )
( )
Figure 12. From left to right (a) epoxy resin mould, (b) posi-
tive model of the part and (c) sintered part.
Figure 14. Drying of alumina cylinder (9 mm diameter with
L/D = 4) in aqueous and non-aqueous solution of PEG1000.
Figure 13. Left to right (a) tube, (b) screw and (c) machined
complex alumina parts formed by gelation of acrylamide.
Forming Alumina Parts Using Acrylamide GelsBarati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003) 135
itive of the part (Figure 12).
FormingAs mentioned earlier, the fabrication of complex-shaped ceramic parts has been limited to processesrequiring extensive and expensive machining; alsothese existing forming processes have inherent prob-lems. For example, one of which is the fabrication ofparts containing thick and thin sections. In this study,many simple and complicated parts have been formedby gel-casting. Figure 13 shows some parts produced inthis work, which indicates the capability of this processfor forming both simple and complex shapes.
DryingConventional methods in drying ceramic parts, usingdry air or air with particular conditions of humidity and
temperature usually confront with many difficulties.Non-uniform and differential drying in various regionsdue to the solvent gradient, induces structural andresidual stresses. They cause defects such as cracking,warping, bending and other malformations, making theparts useless. In this work, a novel method has beendeveloped for drying gel-cast ceramic parts, which notonly increased the drying rate but also removed all ofthe defects due to the release of residual stresses [18].
In this method, by using liquid desiccant (aqueousor alcohol solution of PEG 1000), the parts loose safe-ly about 20-30 wt% of interior solvent (water) in a veryshort time, i.e. 2 h, (Figure 14). It is really a break-through in this critical stage of the process. The linearshrinkage of the part is nearly uniform (Figures 15 and16). The liquid desiccant can provide buoyant supportfor the article being dried. Such support can be helpful
6
5
4
3
2
1
0
Time (min)
Length
Diameter
0 50 100 150 200 250 300
( )( )
Figure 15. Linear shrinkage of alumina cylinder in aqueous
solution of PEG1000.
Time (min)
0
1
2
3
4
5
6
0 30 60 90 120 150 180 210
Length
Diameter
( )( )
Figure 16. Linear shrinkage of alumina cylinder in non-aque-
ous solution of PEG1000.
Figure 18. From left to right, (a) gel-cast alumina part
(35 vol %) dried in room condition, (b) same part dried in
aqueous solution of PEG1000.
Figure 17. From left to right (a) acrylamide pure gel-dried in
non-aqueous solution of PEG1000, (b) same gel dried in air
at room condition.
136
Forming Alumina Parts Using Acrylamide Gels Barati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003)
in preventing distortion of the article during drying(Figures 17 and 18). After the part passes the criticalstage of drying safely, it can continue to dry in the airat room condition or in an oven at higher temperaturewithout any danger (Figure 19).
MachiningMachining of a green part is a cost effective techniquesince the method does not require special hard tools asfor the machining of sintered ceramics. The machin-ability of a green parts depends on various factors suchas monomer concentration, monomer to cross-linkerratio, average particle size, rotation velocity of partaround the axis of machine (rotational velocity of spin-dle), forwarding velocity of tools on the part (feeding
rate), water content of the part and amount of plasticiz-er which was added to the slurry before casting. Basedon the results of Janney et al. [19], experiments weredesigned as in the following details.
As shown in Table 4, lower concentration ofmonomer and monomer to cross-linker ratio cause larg-er chips on the edge and corner of part during machin-ing. This defect can be minimized by using plasticizeror machining the wet green part (with small water con-tent therein). The amount of plasticizer or water content
PEG 20%
PEG 40%
Oven 60OC
Oven 80OC
0 60 120 180 240 300 360 420
100
90
80
70
60
50
40
30
20
10
0
Time (min)
Figure 19. Drying of alumina gel-cast part in various concen-
tration of non-aqueous solution of PEG1000 and in oven at
60 and 80°C.
Figure 20. SEM Photograph of green alumina gel-cast part
(50 vol.% loading in slurry). Note the binder networks holding
the alumina particles together, as indicated by arrows.
Figure 22. Failure of alumina gel-cast part after sintering
at 1450°C because of knit lines therein.
Figure 21. Comparison of dimension and density in (a) cylin-
drical alumina part dried in aqueous liquid desiccant solution
(20 wt %) and in oven at 100°C, (b) after debinding at 600°C
and rate of 2°C/min.
17.2
(a)
4.3
Density = 1.68 Density = 1.68 Density = 1.68 Density = 1.688.7
17.2
(b)
4.3
Density = 1.64 Density = 1.64 Density = 1.64 Density = 1.648.7
of part has a limiting value. Above this level, plasticdeformation occurred on the corner or edge of the part.Higher rotational velocity of spindle and feeding rate,increased the chips on the corner and edge of the spec-imen part. Type of ceramic had no effect on themachining conditions of the green parts.
DebindingAs mentioned earlier, in the gel-casting system ceram-ic particles bind together by using polymeric network(Figure 20). For achieving higher density, this poly-meric binder should be removed from the green part,i.e. oxidation at higher temperature. The pre-mix solu-tions, which are used in this work, contain a maximum10 vol.% polymeric binder (Table 2). As a result, thedry green parts contain only 4 wt% binder that is oftenlower than the binder content in commercial powdersfor compaction. Unlike polymer- or wax-based injec-tion moulding, binder removal is neither time consum-ing, nor critical. STA Analysis of pure acrylamide gelsindicates that oxidation of the acrylamide binder startsat 350-400 C [20]. Above this temperature the heating
rate has to be carefully controlled to avoid build up ofexcess internal pressure (i.e., 0.1 C/min). As seen inFigure 21, there is no shrinkage in the part duringdebinding.
SinteringAfter sintering, knit lines and air bubbles, which are themajor defects, were found in the gel-cast parts [6]. Theknit lines were formed during the gelation stagebecause the curvature and orientation of lines suggest-ed them (Figure 22). This defect was not observedwhen slurry was gently agitated after casting andbefore gelation, and also when long idle time was used.
The other defect, inherent to solvent-based process-es, is entrapped air bubbles, which is tried to beremoved by using an ultrasonic shaker. A typical sin-tering scheme was shown in Figure 23. Figure 24shows the SEM photograph of fracture surface of gel-cast alumina part sintered at 1450 C.
The densities of sintered alumina parts were meas-ured by Archimedes method (Table 5). All the meas-urements are repeated to check for reproducibility andaveraging. During sintering the shrinkage of the partwas isotropic and averaged 8% to 9% linearly for aslurry containing 40 vol % of the alumina powders.These results indicate that the gel-casting is a near-netshape forming technology with predictable dimension.
CONCLUSION
A generic near-net shape forming process based on thein-situ polymerization and gelation of acrylamidemonomer binder, has been evaluated using commercialalumina powders with specific characteristics. Theprocess was found to have the following features:(i) The alumina slurry has low viscosity at high solidsloading (50 vol.% or >80 wt %) by the aid of polyelec-trolytes and TEMED as an accelerator in polymeriza-tion reaction. (ii) The forming process is capable of producing near-
Forming Alumina Parts Using Acrylamide GelsBarati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003) 137
Characteristics Fired at 1450oC
and holder for 1 h
Fired at 1450oC
and holder for 2 h
Bulk density (g/cm3)
Open porosity (%)
Relative density (%)
Close porosity (%)
2.4732
28.37
62.14
9.49
2.6077
27.72
65.54
6.74
Table 5. Results of Archimedes test for alumina parts after
sintering at 1450oC.
1400
1200
1000
800
600
400
200
00 500 1000 1500 2000 2500
Time (min)
slope = 0.375
slope = 2 slope = 2
Figure 23. Typical sintering scheme.
Figure 24. SEM Photograph of fracture surface of gel-cast
alumina part sintered at 1450°C.
138
Forming Alumina Parts Using Acrylamide Gels Barati A. et al.
Iranian Polymer Journal / Volume 12 Number 2 (2003)
net complex shapes.(iii) The time available for casting the slurry (idle time)can be controlled by the amounts of initiator and cata-lyst added to the slurry as well as by the compositionand processing condition such as monomer concentra-tion, monomer to cross-linker ratio, ceramic loadinglevel and temperature.(iv) By using liquid desiccant drying method, the partslose safely about 20-30 wt.% of interior solvent (water)in a very short time without creating any defect ordeformation in them.(v) The green dried gel-cast parts are sufficiently strongfor machining operation. The machinability of a greenpart depends on factors such as monomer concentra-tion, monomer to cross-linker ratio, average particlesize, rotation velocity of the part around the axis ofmachine, forwarding velocity of the tools on the part,water content of the part and amount of plasticizer. (vi) The binder concentration is low, less than 4 wt %of the dried solid. Therefore, binder removal is neithertime consuming, nor critical.(vii) The total linear shrinkage of a 40 vol % aluminaslurry was 4% during drying and 8% to 9 % during par-tial sintering.
ACKNOWLEDGEMENTS
The authors wish to thank Tarbiat Modarres University(TMU) for the financial support of this work. Also,thank the co-workers and technical staff in the Depart-ment of Polymer Engineering of the TMU for their helpduring various stages of the work.
REFERENCES
1. Rahaman M.N., Ceramic Processing and Sintering,6, Marcel Dekker, NY, 309 (1995).
2. Wang L. and Aldinger F., Near-net shape forming ofadvanced ceramics, Adv. Eng. Mat., 2, 110 (2000).
3. Reed J.S., Principles of Ceramic Processing 2/e,Chaps. 2 and 3, Wiley Int., NY, USA (1995).
4. Tari G., Ferrera J.M. and Fonseca A.T., Influence ofparticle size and particle size distribution on drying-shrinkage behaviour of alumina slip cast bodies,Ceram. Inter., 25, 577-580 (1999).
5. Edirisinghe M. J., Removal of organic vehicle frommoulded ceramic bodies, Br. Ceram. Proc., 45, 109-122 (1990).
6. Omatete O.O., Janney M.A. and Strehlow R.A., Gel-
casting - A new ceramic forming process, Am.Ceram. Soc. Bull., 10, 1641-1649 (1991).
7. Gilssen R., Erauw J.P., Smolders A., Vanswijgen-hoven E. and Luyton J., Gel-casting, a near-netshape technique, Mat. and Des., 21, 251-257 (2000).
8. Loure J., Bagnall C. M., Harris B., Colloidal suspen-sions for the preparation of ceramics by a freeze cast-ing route, J. Non. Cryst. Solids, 147, 320-325 (1992).
9. Millan A. J., Moreno R. and Nieto M. I., Aqueousinjection moulding of alumina using agarose, Mater.Lett., 47, 324-328 (2001).
10. Si W., Groule T.J., Baader F.H. and Gauckler L.J.,Direct coagulation casting of silicon carbide compo-nents, J. Am. Ceram. Soc., 82, 1129-1136 (1999).
11. Kosmac T., The Potential of the hydrolysis assistedsolidification (HAS) process for wet forming of engi-neering ceramics, Key Eng. Mater., 161, 357-362 (1999).
12. Young A.C., Omatete O.O., Janney M.A. andMenchhofer P.A., Gel-casting of alumina, J. Am.Ceram. Soc., 74, 612-618 (1991).
13. Kokabi M., Pirooz A. and Nekoomanesh HaghighiM., Gel-casting of engineering ceramics, Iran. Pol.J., 7, 169-175 (1998).
14. Barati A., Kokabi M. and Babaluoa A.A., Effect ofpoly(methacrylic acid) polyelectrolytes in stabilityof alumina suspensionsin, Persian, Iran. Polym. J.Sci. Tech., 2, 131-141 (2002).
15. Lem K. and Han C.D., Chemorheology of the ther-mosetting resins. II. Effect of particulates on thechemorheology and curing kinetics of unsaturatedpolyester resin, J. Appl. Polym. Sci., 28, 3185-3206(1983).
16. Ng H. and Mansas-Zlockzower I., Chemorheologyof unfilled and filled epoxy resins, Polym. Eng. Sci.,33, 4, 211-216 (1993).
17. Paauw M. and Pizzi A., Some filler affects on cross-linking of unsaturated polyester, J. Appl. Polym. Sci.,50, 1287-1293 (1993).
18. Barati A., Kokabi M. and Famili N., Modelling ofliquid desiccant drying method for gelcast ceramicparts, Ceram. Inter., 29, 199-207 (2003).
19. Janney M.A. and Walls C.A.H., Gel-casting com-positions having improved drying characteristics andmachinability, US Patent, 6228299 B1, May 2001.
20.Tahooni N. and Kokabi M., Study on gelation mech-anism of acrylamide system and its thermal behav-iour, in Proceeding of the 3rd National Conf. onChemical Engineering, Ahwaz, Ed. R. Kharrat,Petroleum Industry University Pub., Ahwaz, I.R.Iran) p 384 (Mar. 1998).
Top Related